Data-Driven Prescriptive Analytics Applications: A Comprehensive Survey

TL;DR

This survey defines Data-Driven Prescriptive Analytics (DPSA) as end-to-end, automated workflows that combine data-driven prediction with automatic prescription and analyzes 104 DPSA papers to map problem domains, methods, and workflow patterns. It introduces a taxonomy of 10 application domains, 5 method types, and 2 generic DPSA workflow patterns, plus 5 strategic directions for future work. The analysis shows a strong emphasis on Data Mining/ML for prediction and Mathematical Optimization for prescription, with substantial but evolving use of Probabilistic Modelling, Domain Expertise, and Simulation; two generic workflow patterns (PTP and PWP) structure most DPSA approaches. The paper also highlights under-explored domains, the need for scalable alternatives to (mixed-)integer linear programming, and broader tooling and production deployment practices to advance DPSA adoption in industry.

Abstract

Prescriptive Analytics (PSA), an emerging business analytics field suggesting concrete options for solving business problems, has seen an increasing amount of interest after more than a decade of multidisciplinary research. This paper is a comprehensive survey of existing applications within PSA in terms of their use cases, methodologies, and possible future research directions. To ensure a manageable scope, we focus on PSA applications that develop data-driven, automatic workflows, i.e., Data-Driven PSA (DPSA). Following a systematic methodology, we identify and include 104 papers in our survey. As our key contributions, we derive a number of novel taxonomies of the field and use them to analyse the field's temporal development. In terms of use cases, we derive 10 application domains for DPSA, from Healthcare to Manufacturing, and subsumed problem types within each. In terms of individual method usage, we derive 5 method types and map them to a comprehensive taxonomy of method usage within DPSA applications, covering mathematical optimization, data mining and machine learning, probabilistic modelling, domain expertise, as well as simulations. As for combined method usage, we provide a statistical overview of how different method usage combinations are distributed and derive 2 generic workflow patterns along with subsumed workflow patterns, combining methods by either sequential or simultaneous relationships. Finally, we derive 5 possible research directions based on frequently recurring issues among surveyed papers, suggesting new frontiers in terms of methods, tools, and use cases.

Figures (14)

• Figure 1: An Analytics Ascendancy Model, adapted and modified for this paper from siksnys2018prescriptive. Modifications include the omission of Diagnostic Analytics as a separate area from Descriptive Analytics (cf. lepenioti2019prescriptive), preferring Complexity over the more subjective Difficulty axis, as well as core question wordings.
• Figure 2: Flow diagram of the executed survey procedure, using the PRISMA 2020 format page2021prisma. Each box in the left chain represents a phase in a sequential screening workflow, along with its number of input records. The boxes to the right summarize how many records were excluded in each screening phase, as well as the reasons why, with respect to the selection criteria of Section \ref{['sc:method:criteria']}.
• Figure 3: Number of papers per year in the set of surveyed papers.
• Figure 4: Stacked bar chart illustrating the number of papers published within each problem domain over time, as well as the distribution of published papers per year.
• Figure 5: Bar chart summarizing the relative frequency of the different method types in the set of surveyed papers. Note that this aggregation does not form distinctions based on the role played by each method on the DPSA workflow in terms of providing predictions or prescriptions.

Theorems & Definitions (3)

• definition 1: Business Analytics
• definition 2: Prescriptive Analytics
• definition 3: Data-Driven Prescriptive Analytics